I picture copper tubing coiled around the the duct not too far down stream from the manifold. Is that like what you had in mind?
I'm developing a physical computing
class for the middle school where I work. It seems like this would be an interesting place to aim some adult lesson plans. I can imagine a combined workshop with participants learning the science of the
RMH, building one, and embedding a sensor network to measure things like the content and temperature of the exhaust gas along the length of the duct. Then, once it's cured, someone could burn the thing for a long time and keep track of the internal conditions of the system. That would help to decide where to place the coil in the final design to make as sure as possible it will not be able to heat water beyond a safe temperature.
You could also embed temp sensors in the mass to help locate a different place to locate some kind of mass-to-water heat exchanger. That would be easy in a pebble style
RMH - just dig up the sensors and burry the exchanger in the same place. With
cob and the like you'd have to build a test bench before the final bench or design for partial demolition and repair
Since I'm kind of a geek I would also want to incorporate thermoelectric generation to run the controller and maybe even the pump - and charge this laptop when the pump isn't running, and maybe an
LED lamp for when I'm up late reading Permies.
The design would need to protect the unit(s) from excessive heat and make them easily accessible because I'd want to experiment with them.
I don't think this kind of high tech stuff turns up often on the forums so I hope it doesn't offend anyone's sensibilities. But this stuff has become so cheap and accessible, I've come to view them as an
appropriate technology to apply in the development and refinement of low-tech. We use electrically powered vehicles and tools in our large-scale projects, why not an electrically powered (and cheap) nervous system in our smaller-scale projects?